Smart type casting in python

Question

I am making api calls and receive a response back in json like so

result = {'foo': '123.456', 'bar': '23', 'donald': 'trump', 'time': '2016-04-07T05:31:49.532124Z'}

Although result is either a dictionary or a list, the contents are always strings. I want to cast these values to the appropriate type. (ie '123.456' to a float, '23' to an int, 'time' to a datetime.)

My code works but it seems like there should be a simpler/more efficient way to do this. Is there?

Here's my version

from dateutil.parser import parse

def is_float(x):
    try:
        float(x)
        return True
    except ValueError:
        return False

def is_int(x):
    try:
        a = float(x)
        b = int(a)
    except ValueError:
        return False
    else:
        return a == b

def is_datetime(x):
    try:
        parse(x)
        return True
    except ValueError:
        return False

def smartcast(x):
    if isinstance(x, dict):
        return { k:smartcast(v) for k, v in x.items() }
    elif isinstance(x, list):
        return map(smartcast, x)
    elif is_int(x):
        return int(x)
    elif is_float(x):
        return float(x)
    elif is_datetime(x):
        return parse(x)
    else:
        return x

Edit: If possible I'd like to avoid using try and except . I'd like to implement this into an asynchronous twisted program and I think try and except block so it just makes the program synchronous. I'm new to twisted so I'm not sure if that is true.

---

As per Francesco's solution, here's the updated code.

from dateutil.parser import parse

def smartcast(x):
    if isinstance(x, dict):
        return { k:smartcast(v) for k, v in x.items() }
    elif isinstance(x, list):
        return map(smartcast, x)
    else:
        for t in [int, float, parse]:
            try:
                return t(x)
            except ValueError:
                continue
    return x

Answer 1

And what is the appropriate type ? I am asking this to underline that sometimes also for humans it is not clear what the type is: are you sure that "42" is always an int or sometimes it has to be considered just as a string?

Anyway, as you did, you can build your solution with your rules that makes sense in your context.

In your case, you can simplify your code with a simple loop

from dateutil.parser import parse

tests = [int, float, parse]

def smartcast(value):
     for test in tests:
         try:
             return test(value)
         except ValueError:
             continue
     # No match
     return value

Answer 2

what about using ast.literal_eval()?

https://docs.python.org/2/library/ast.html#ast.literal_eval

voting up Francesco's answer. You need an exception raised.